I also ran this on home servers, for a couple of years (2022-2023). It was certainly performant, but I migrated off when I realized that it was too "opinionated" to gain my full trust. Updates broke it badly several times. I think the article citing its "promise" is accurate; in my case I was looking for an offramp for my home/personal use of CentOS. When Red Hat announced their developer allowance of up to 16 RHEL hosts for free, that crystallized the cost of the amount of toil I expended in trying to maintain Clear Linux.

With the right combination of requisite scale and need for performance, it would have been a good choice.

I gather from the reports that they didn't see a path towards moving the product to a community support and development model, which says something.

It's too bad, and I remain PO'd at what happened to CentOS...

This reminds me of the somewhat similar issues around navigation in the physical world, where search engines replace maps - this time, the search engines show you all (only) what you need to know to reach your destination.

Whatâ(TM)s missing? Context, sense of place, understanding of how a set of files relate to each other (same directory) and the ability to think of a route from one place to another without using the mobile map.

Like a search engine that is excellent at answering the (right) question you ask of it, GPS and mapping systems will certainly tell you how to get to the location you programmed - but this assumes the stuff you see along the way is of no importance. Not noticing thereâ(TM)s a notable town just a bit further out of the way, or recalling that one had some data samples and references in the same folder as the code - I guess thatâ(TM)s something one (hopefully) learns from experience, eventually.

All of my devices will "self-destruct" my passwords if I can't type them in a certain number of attempts. And my pass-phrases are so long... and complicated. Not sure I could really do them correctly under duress. My USB as well - same thing, will fry itself after a few bad attempts. I'm just not sure which is for which. Sorry.

I never really looked at this stuff previously, but got to pick up on a project that had been started already. The installers used cameras from AXIS; I'm reasonably impressed by them so far.

It sounds like you're thinking the camera should "push" and you're asking why the end-user can't control "where to". The unit that I have (M1114) gives you the ability to "pull" - this might be a better fit to your needs. Yes, a bit trickier than something that self-publishes, but much more control.

(The big issue here was this is being used for surveillance, and there were many choices for server software, most of which were Windows-based - I'll put in a plug for 'motion', a package that we're running on Ubuntu and has been pretty flawless for what we needed. People here were impressed by the result of both the software and the hardware.)

Don't give up, there are good industrial-grade IP camera solutions out there!

Yep... clearly there's a some confusion about bandwidth v. usage.

My understanding was that Comcast capped my usage (I have a "Blast Plus" account) at 250GB/month. Seems to reset every month at 0000Z on the first.

I monitor my own usage fairly carefully, I think last month I was a bit over 300GB. But: Comcast announced that they are temporarily suspending enforcement of their cap - as of May, 2012. They have indicated that they are going to replace the cap with "new approaches", but there doesn't seem to be any mention of this since that date.

There's clearly tension between the cost of the plant necessary to support large users of data and the profit of the corporation. Without diving into that debate, I will relate what I've heard in discussions with a few other technically-oriented customers: they really didn't engineer their network to support any significant level of symmetric usage (it's designed for some multiples of download traffic over upload traffic), and the whole question of traffic shaping was a response to an actual "fear" that their network would be trashed by BitTorrent users.

I don't know whether that's actually true. But I do know that ultimately, all resources on the Internet are shared, and that without sufficient bandwidth there is lots of potential that won't be attained. I have some level of sympathy for the idea that someone may have to be throttled to keep costs for usage roughly equal for all customers of a particular class, or that pricing might have to be "reasonably" adjusted based on usage - especially peak usage (I do offsite vaulting exclusively between 0100 and 0500 local time).

I'd hope that Comcast and the other ISPs realize that the only way they are going to be able to make their business case to their customers if they operate transparently - and I know that's not the initial strategy for any large corporation. I don't think most people have an issue with paying "reasonable" fees - but, again, in the telecomm/service industry in general, that's generally how things are priced.

They are going to have to figure out how to make some profits on what anyone reading this already considers a "utility" - they sure aren't getting revenue from my pageviews of the ads on their home page. However, I've heard too many stories of how Internet connectivity delivered as a pure utility (especially by non-profits such as municipal governments) provides better service, higher bandwidth (sic) and lower costs than investor-owned corporations; if I could vote (if it were practical in my section of Boston) for that, as a public good, it'd make this discussion go away.

/. is actually the default page when opening Chrome for me - so it's not just that I have one open right now, it's that /. is the first one that opens when I launch.

This may or may not be helpful.

I have my own server running WebDAV. It's "stock" Apache, the DAV module is included. A couple of people have mentioned Subversion, but of course that overlays functionality that may not be necessary or practical. DAV is fairly simple, and dedicated to supporting an essential set of semantics for accessing files.

The reason this may not be useful to you is that my primary use of the server has been as my "remote" fileserver, for Mac and Windows clients. In some sense, it'd be easier if I wanted to set up CIFS (Samba) or AFP (Apple), but neither of these work all that great when you have significant latency, which is the case if the server isn't right there physically to where you're working. Many iOS apps have taken to being written to directly talk to things like Dropbox, or WebDAV. That may not be true about the Android Apps that you care about. I agree with the premise that Dropbox doesn't give a warm fuzzy security feeling, though the functionality is very nice.

Setting up WebDAV on Apache, secured with TLS, isn't all that difficult. The protocol works well in a high-latency environment. The one technical issue that I've found is that there aren't good solutions for access control - it's not easy to set up multiple user accounts for one server and enforce separate access for those users. I've looked at a few kludges but nothing that seems satisfying.

Depending on your specific need, though, don't rule WebDAV out.

Sure, I understand that Microsoft implements SIP and RTP in these products, just like Exchange implements SMTP and IMAP. But they now own the other side of the conversation - meaning that it appears you'll need a Microsoft Windows Server-based gateway to bridge SIP and Skype. This is a different proposition than a $70 licence (per channel/session, I think), which is the current cost of running the Asterisk/Skype gateway. Skype as a closed protocol is now owned by a company whose business model historically is based on closed protocols; Microsoft presumably sees the value of limiting the ability of non-Microsoft PBXs to interoperate with it. Lync Server is probably a fine solution for an environment that wishes to integrate SIP into a Windows/Exchange/AD workplace, but carries costs and risks for those not wanting to have their infrastructure assimilated, er, integrated in this manner. And the direction that Microsoft intends to take with this acquisition isn't yet clear. If Microsoft were to open the Skype protocols or migrate Skype to SIP or some other open standard, I think I'd be cheering.

The Microsoft way of doing things is to keep things closed. Skype is closed, SIP - whether you're using Asterisk or some other solution - is open. The Asterisk product was a good way to provide an organization with a gateway they could control to bridge SIP in the enterprise with Skype. Who cares about Asterisk... if there are other gateways between SIP and Skype? Except there aren't. It *appears* that this is a way to keep what's closed closed. That's the underlying pain here.